🔍 Read the full analysis: AI Agents Are Opening GitHub PRs At Scale—Can Review Keep Pace? on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
GitHub’s Octoverse 2026 report says more than 40% of pull requests in September began with a coding agent’s first draft, up from 17% in March. The company also introduced AI review and model-routing tools as it confronts the human review burden, but its figures do not show whether merged agent code is correct.
More than 40% of pull requests on GitHub in September began with a coding agent’s first draft, up from 17% in March, according to GitHub’s Octoverse 2026 report, released October 28 at its Universe conference. The company says that amounted to 110 million agent-started pull requests in one month, a surge that puts pressure on people and tools responsible for reviewing proposed code.
GitHub defines an agent-started pull request as one whose first draft came from a coding agent the platform can detect. It says the figure is a floor because some agent use may go undetected. The metric includes public and private repositories, according to the report’s methodology, but GitHub did not provide a separate breakdown for open-source and company code. A pull request can still count as agent-started if a person later substantially rewrites the draft.
GitHub reports that 90% of agent pull requests were merged within two weeks in September, compared with 83% in February. The reported rate for pull requests without a detected agent was about 86%. Those figures measure whether changes were merged, not whether the code was correct, maintainable or free of later defects.
At Universe, GitHub introduced Copilot code review Max, which brings together reviews from several AI models, and Copilot Fusion, a system that routes tasks to models. The company also made Copilot review approvals eligible to count toward required approvals. GitHub estimates an average review costs $0.25 for Lite, $1.50 for Balanced and $10 for Max, with GitHub Actions minutes charged separately.
Infographic 01 · The agent share
Four in ten pull requests now start with an agent.
GitHub’s own telemetry, March to September 2026. The first draft comes from an agent; review and merge stay with people.
Merged within two weeks
Share of pull requests merged within 14 days
Merged fast is not the same as correct. GitHub publishes no revert or defect rate.
Bigger changes to review
Share of pull requests that change more than one file
More files per change means more for a reviewer to check.
What GitHub counts: an “agent-started pull request” is one “whose first draft came from a coding agent that we can detect.” Undetected agents are missing, so GitHub calls the figure a floor. Metrics cover public and private repositories unless noted.
Source: GitHub Octoverse 2026, 28 Oct 2026 (text, chart captions, glossary and methodology). All figures are GitHub’s own.
Powered by Thorsten Meyer AI
Review Becomes the New Bottleneck
The figures point to a shift in where software teams may spend time. If agents produce more initial drafts, the amount of code requiring human attention, testing and approval can grow even when drafting gets faster. GitHub’s launch of review products alongside its agent figures reflects that operational challenge: producing code is only one stage in getting a change safely into a product.
For organizations, the reported prices make review capacity a budget question as well as a workflow question. Teams will need to decide when an AI review is useful, whether its approval can satisfy internal rules, and how much human review remains necessary. Counting pull requests or merges alone cannot establish that agents improve software quality; measures such as rework, reversions and production incidents would provide additional evidence.
For developers, the change may increase the value of writing clear task instructions and critically examining diffs. It could also affect how new engineers learn: small, bounded coding tasks have often served as practice, while those are among the tasks agents can take on. The report does not quantify effects on hiring or training, so those implications remain possible consequences rather than established findings.
Products Worth Considering
Logitech MX Mechanical Wireless Illuminated Keyboard Tactile (Graphite)
Dell UltraSharp U2725QE 27-Inch 4K UHD IPS Black Monitor
Dell UltraSharp 27 Thunderbolt Hub Monitor (U2726DER)
As an affiliate, we earn on qualifying purchases.
How GitHub Measures Agent Work
Octoverse is GitHub’s yearly report on activity across its platform. The 2026 edition covers October 1, 2025, through September 30, 2026. GitHub says the agent share rose from 17% in March to more than 40% in September, and describes the September total as about six times March’s volume. The platform’s definition relies on agent activity it can detect, not a complete census of every tool used to write code.
The report also describes broader platform growth: GitHub recorded 1.3 billion merged pull requests during the reporting year, compared with 402 million in 2024 and 538 million in 2025. Those earlier figures were recalculated using updated definitions and a revised reporting window, so they should not be treated as directly comparable with older editions without that qualification. GitHub says agent use increased as tools moved toward persistent, cloud-based work, but the report does not isolate how much of overall platform growth agents caused.
GitHub’s wording varies across the report, including “agent-started,” “agent-generated” and “agent-opened.” Its formal definition is based on a first draft from a detectable coding agent. That narrower wording is the clearest description of what the headline percentage measures.
“A pull request whose first draft came from a coding agent that we can detect.”
— GitHub’s Octoverse 2026 report
Quality and Accountability Stay Open
The report does not establish whether agent-started changes are more or less reliable than human-started ones. Merge rates are not quality measures, and the source material does not provide comparative data on defects, security issues, rework or long-term maintenance. It also does not say how many of the 110 million September pull requests were later changed, reverted or linked to production incidents.
Other limits include the lack of a public-versus-private repository split and the possibility that undetected agent activity is missing from the count. GitHub says its cloud agent cannot approve or merge its own pull requests, and that the person who requested a change cannot approve it. The report does not detail how every organization will apply those safeguards or what responsibility rules apply across different team policies.
Teams Must Test Review Policies
The next test will be whether GitHub’s review and routing features help teams handle larger volumes without weakening their standards. Organizations can compare the tools’ costs with reviewer time and track rework, reversions and incidents alongside merge volume. The source material does not give a schedule for adoption data or independent evaluations of the new review products.
GitHub’s October report provides a snapshot through September 30, while its Universe launches introduce tools intended to address review. Further evidence will be needed to show how widely those features are used, how often AI approvals count in real workflows, and whether code quality changes as agent-started pull requests become more common.
Key Questions
What does GitHub count as an agent-started pull request?
GitHub counts a pull request when its first draft came from a coding agent the platform can detect. Human edits made afterward do not necessarily remove it from the count.
Does the 40% figure mean 40% of GitHub code is AI-written?
No. It refers to the share of September pull requests that GitHub classified as starting with an agent’s first draft. It does not measure the share of all code written by AI or how much of each pull request an agent produced.
Were agent-started pull requests more likely to be merged?
GitHub says 90% were merged within two weeks in September, versus about 86% for pull requests without a detected agent. A merge does not establish that code was correct or defect-free.
How much do GitHub’s AI reviews cost?
GitHub estimates average costs of $0.25 for Lite, $1.50 for Balanced and $10 for Max review. GitHub Actions minutes are additional, and actual costs may depend on use.
Who is responsible when agent-written code ships?
GitHub says developers remain accountable for shipped code. It also says its cloud agent cannot approve or merge its own pull requests, and the person who requested the change cannot approve it.
Source: ThorstenMeyerAI.com
